Competitive carbon-sulfur vs carbon-carbon bond activation of 2-cyanothiophene with [Ni(dippe)H]2

Article abstract of DOI:10.1021/ja104158h

The processes of C-C and C-S bond cleavage have been studied with the homogeneous organometallic compound [Ni(dippe)H]₂ (1). When 1 is reacted with 2-cyanothiophene at room temperature, cleavage of the nitrile-substituted C-S bond occurs, forming the Ni-metallacycle complex (dippe)Ni(κ²-S,C-SCH=CHCH=C(CN)) (2a), which has been fully characterized by NMR spectroscopy and X-ray diffraction. 2a was converted to the C-CN cleavage product (dippe)Ni(CN)(2-thiophenyl) (3) when heated in solution. On closer inspection, four other intermediates were observed by ³¹P NMR spectroscopy at low temperature. Structures for the intermediates were elucidated through a combination of independent synthesis, theoretical calculations, chemical characterization, and experimental precedent. A kinetic product (dippe)Ni(κ²-S,C-SC(CN)=CHCH=CH) (2b) was formed from cleavage of the nonsubstituted C-S bond, as well as a Ni(0) η²- nitrile intermediate, (dippe)Ni(η²-C,N-2-cyanothiophene) (4), and a dinuclear mixed Ni(0)-Ni(II) product (6b). A complete DFT analysis of this system has been carried out to reveal comparative details about the two bond cleavage transition states.

Full text of DOI:10.1021/ja104158h

Published on Web 08/12/2010
Competitive Carbon-Sulfur vs Carbon-Carbon Bond
Activation of 2-Cyanothiophene with [Ni(dippe)H]2
Matthew R. Grochowski, Ting Li, William W. Brennessel, and William D. Jones*
Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627
Received May 14, 2010; E-mail: jones@chem.rochester.edu
Abstract: The processes of C-C and C-S bond cleavage have been studied with the homogeneous
organometallic compound [Ni(dippe)H]
2
(1). When 1 is reacted with 2-cyanothiophene at room temperature,
2
cleavage of the nitrile-substituted C-S bond occurs, forming the Ni-metallacycle complex (dippe)Ni(κ -
S,C-SCHdCHCHdC(CN)) (2a), which has been fully characterized by NMR spectroscopy and X-ray
diffraction. 2a was converted to the C-CN cleavage product (dippe)Ni(CN)(2-thiophenyl) (3) when heated
in solution. On closer inspection, four other intermediates were observed by 31P NMR spectroscopy at low
temperature. Structures for the intermediates were elucidated through a combination of independent
synthesis, theoretical calculations, chemical characterization, and experimental precedent. A kinetic product
2
(
dippe)Ni(κ -S,C-SC(CN))CHCHdCH) (2b) was formed from cleavage of the nonsubstituted C-S bond,
2
2
as well as a Ni(0) η -nitrile intermediate, (dippe)Ni(η -C,N-2-cyanothiophene) (4), and a dinuclear mixed
Ni(0)-Ni(II) product (6b). A complete DFT analysis of this system has been carried out to reveal comparative
details about the two bond cleavage transition states.
Introduction
is employed, Ni/Mo (or Mo/Co) sulfide, is effective for
removing most organo-sulfur compounds. However, removal
The processing of petroleum represents perhaps the largest
scale organic chemical process currently in practice. Two of
the most important steps in this processing are catalytic cracking
of substituted thiophenes remains problematic, and improvement
in catalyst efficiency will require increased effectiveness toward
4
these compounds.
(
C-C bond cleavage) and hydrodesulfurization (C-S bond
1
Transition metal compounds that cleave C-C bonds in
homogeneous solution are relatively limited. General examples
that have been established involve energetically favorable
cleavage). The petroleum industry employs large-scale hetero-
2
geneous (zeolitic) catalysts for effecting these reactions. Over
the past decades, homogeneous transition metal compounds have
been studied with the goal of providing fundamental information
about these bond-breaking reactions.
Fluid catalytic cracking (FCC) is the most important conver-
sion process used in petroleum refineries. It is widely employed
to convert the high-boiling, high-molecular-weight hydrocarbon
fractions of petroleum crude oils to more valuable gasoline,
olefinic gases, and other products. The zeolite used in FCC
catalysts is referred to as faujasite or Type Y, and the catalytic
5
reactions that depend on relief of ring strain, attainment of
6
aromaticity, or reduction of entropic factors via intramolecular
7
C-C cleavage. Examples of alkane C-C cleavage by homo-
geneous metal complexes are rare, although there have been
recent reports that take advantage of dehydrogenation/metathesis
8
9
pathways or metal radical pathways to achieve aliphatic C-C
2
cleavage. Our laboratory has reported sp-sp C-C cleavage
in aliphatic, aromatic, and allylic nitriles. In these ex-
10
11
12
3
2
amples, the complex [Ni(dippe)H] (1) reacts to form complexes
sites are strongly acidic.
2
in which the [Ni(dippe)] fragment first coordinates η to the
nitrile Ct N (R ) alkyl or aryl) or CdC (R ) allyl, aryl) moiety.
Rearrangement then occurs to give the C-CN oxidative addition
Hydrodesulfurization (HDS) is the process by which sulfur
is removed from hydrocarbons during the refinement of
petroleum. Failure to remove sulfur during this process results
in the formation of noxious sulfur oxides upon fuel combustion
which are a major cause of air pollution. Current catalyst
efficiency must be improved to meet increasingly rigorous
environmental standards. The current heterogeneous catalyst that
(
4) Topsoe, H.; Clausen, B. S.; Massoth, F. E. Hydrotreating Catalysis;
Springer: Berlin, 1996.
(5) Adams, D. M.; Chatt, J.; Guy, R. G.; Sheppard, N. J. Chem. Soc.
961, 738–742. Steffen, A.; Ward, R. M.; Jones, W. D.; Marder, T. B.
1
Coord. Chem. ReV. 2010, 254,in press.
6) Benfield, F. W. C.; Green, M. L. H. J. Chem. Soc., Dalton Trans.
1974, 1324.
(
(
1) Gary, J. H.; Handwerk, G. E.; Kaiser, M. E. Petroleum Refining
Technology and Economics, 5th ed.; CRC Press: Boca Raton, 2007.
Leffler, W. L. Petroleum Refining for the Nontechnical Person, 2nd
ed.; PennWell Books: Tulsa, OK, 1985.
(7) Suggs, J. W.; Jun, C.-H. J. Chem. Soc., Chem. Commun. 1985, 92.
Gozin, M.; Aizenberg, M.; Liou, S.-Y.; Weisman, A.; Ben-David, Y.;
Milstein, D. Nature 1994, 370, 42.
(
(
2) Yang, W.-C. Handbook of Fluidization and Fluid Particle Systems;
CRC Press: Boca Raton, 2003.
3) (a) Gary, J. H.; Handwerk, G. E. Petroleum Refining: Technology and
Economics, 4th ed.; CRC Press: Boca Raton, 2001. Speight, J. G. The
Chemistry and Technology of Petroleum, 4th ed.; CRC Press: Boca
Raton, 2006. Sadeghbeigi, R. Fluid Catalytic Cracking Handbook,
(8) Goldman, A. S.; Roy, A. H.; Huang, Z.; Ahuja, R.; Schinski, W.;
Brookhart, M. Science 2006, 312, 257.
(9) Chan, Y. W.; Chan, K. S. J. Am. Chem. Soc. 2010, 132, 6920.
(10) (a) Garc ´ı a, J. J.; Ar e´ valo, A.; Brunkan, N. M.; Jones, W. D.
Organometallics 2004, 23, 3997. (b) Ate s¸ in, T. A.; Li, T.; Lachaize,
S.; Brennessel, W. W.; Garc ´ı a, J. J.; Jones, W. D. J. Am. Chem. Soc.
2007, 129, 7562.
2
nd ed.; Gulf Publishing: , 2000.
1
2412 9 J. AM. CHEM. SOC. 2010, 132, 12412–12421
10.1021/ja104158h  2010 American Chemical Society

C-S vs C-C Bond Cleavage in 2-Cyanothiophene with [Ni(dippe)H]
2
A R T I C L E S
product Ni(dippe)(R)(CN). One case studied in detail was the
complex if it is left in THF solution (eq 3), indicating reVersible
cleavage of the C-CN bond in benzonitrile, which occurs via
C-S bond cleavage. Reaction of 1 with dibenzothiophene gives
2
2
reversible rearrangement to an η -C,C complex prior to C-C
the analogous (dippe)Ni(κ -C,S-C12
H
8
S)], which undergoes
1
9b
cleavage (eq 1).
desulfurization to give four unique products.
This work specifically addresses the reaction of 1 with
-cyanothiophene, where the possibility for C-CN cleavage
2
Many homogeneous organometallic systems have also been
studied in an attempt to better understand the mechanism(s) of
exists in competition with the various C-S cleavage reactions.
Reaction of 2-cyanothiophene with the platinum analogue
13
14
15
HDS, including model studies of C-S cleavage at Mo, Co,
20
[
Pt(dippe)] has already been examined. The Pt(dippe) fragment
1
6
17
18
Rh, Ir, and Pt. The complex [Ni(dippe)H]
2
(1) has also
was found to cleave both the nitrile-substituted and unsubstituted
C-S bonds. Cleavage away from the cyano group was preferred
kinetically, while cleavage adjacent to the cyano group was
thermodynamically preferred (eq 4). No C-CN cleavage was
been found to insert into the C-S bonds of a wide range of
thiophenes, including the challenging dibenzothiophene, 4-me-
thyldibenzothiophene, and 4,6-dimethyldibenzothiophene sub-
1
9
strates. The reaction of thiophene with 1 is shown in eq 2,
2
where the product (dippe)Ni(κ -C,S-C
4
H
4
S) is typical of the
C-S insertion that takes place. 1 undergoes loss of H
2
when
observed, even after heating the sample at 160 °C for several
months. The reaction was also studied using density functional
theory (B3LYP), and the theoretical calculations were found to
be useful for comparing transition states for the two types of
bond cleavage. The work presented here compares experimental
and computational results, and shows marked differences with
the platinum system.
reacted with a substrate to produce the highly reactive 14-
electron Ni(0)-dippe fragment bound to the substrate, which then
8
19b
continues on to a Ni(II) d square planar product. Interest-
2
ingly, while (dippe)Ni(κ -C,S-C
4 4
H S) is stable under nitrogen
in the solid state, it forms an equilibrium with a dinuclear
(
(
11) (a) Garcia, J. J.; Jones, W. D. Organometallics 2000, 19, 5544. (b)
Garcia, J. J.; Brunkan, N. M.; Jones, W. D. J. Am. Chem. Soc. 2002,
Results and Discussion
Reaction of 1 with 2-Cyanothiophene. The reaction of 1 with
2-cyanothiophene occurred cleanly and rapidly at room tem-
perature in THF to give the C-S insertion complex (dip-
1
24, 9547. (c) Atesin, T. A.; Li, T.; Lachaize, S.; Brennessel, W. W.;
Garc ´ı a, J. J.; Jones, W. D. Organometallics 2008, 27, 3811. (d) Li,
T.; Garc ´ı a, J. J.; Brennessel, W. W.; Jones, W. D. Organometallics
2
010, 29, 2430.
2
pe)Ni(κ -S,C-SCHdCHCHdC(CN)) (2a) in 93% isolated yield
12) Acosta-Ram ´ı rez, A.; Mu n˜ oz-Hern a´ ndez, M.; Jones, W. D.; Garc ´ı a,
J. J. J. Organomet. Chem. 2006, 691, 3895. Acosta-Ram ´ı rez, A.;
Flores-Gaspar, A.; Mu n˜ oz-Hern a´ ndez, M.; Ar e´ valo, A.; Jones, W. D.;
Garc ´ı a, J. J. Organometallics 2007, 26, 1712. Acosta-Ram ´ı rez, A.;
Mu n˜ oz-Hern a´ ndez, M.; Jones, W. D.; Garc ´ı a, J. J. Organometallics
(eq 5). The insertion occurred exclusively at the substituted C-S
´
2007, 26, 5766. Acosta-Ram ´ı rez, A.; Flores-Alamo, M.; Jones, W. D.;
Garc ´ı a, J. J. Organometallics 2008, 27, 1834. Acosta-Ram ´ı rez, A.;
Morales-Morales, D.; Serrano-Becerra, J. M.; Ar e´ valo, A.; Jones,
W. D.; Garc ´ı a, J. J. J. Mol. Catal. A, Chem. 2008, 288, 14. Swartz,
B. D.; Reinartz, N. M.; Garcia, J. J.; Jones, W. D. J. Am. Chem. Soc.
2
008, 130, 8548. See also: Van der Vlugt, J. I.; Hewat, A. C.; Neto,
bond. The 31P NMR spectrum showed two closely spaced
S.; Sablong, R.; Mills, A. M.; Lutz, M.; Spek, A. L.; M u¨ ller, C.; Vogt,
D. AdV. Synth. Catal. 2004, 346, 993. Wilting, J.; Muller, C.; Hewat,
A. C.; Ellis, D. D.; Tooke, D. M.; Spek, A. L.; Vogt, D. Organome-
tallics 2005, 24, 13. Chaumonnot, A.; Lamy, F.; Sabo-Etienne, S.;
Donnadieu, B.; Chaudret, B.; Barthelat, J. C.; Galland, J. C. Orga-
nometallics 2004, 23, 3363.
doublets at δ 73.06 and 72.22 that resemble a virtual quartet
(JP-P ) 34 Hz). Orange block crystals suitable for X-ray
diffraction were grown from a layered solution of benzene/
hexanes at -30 °C. The crystal structure in Figure 1 shows a
nickel metallacycle product with a square planar geometry
around the Ni(II) center. The bond lengths and angles are very
(
13) Sanchez-Delgado, R. A. Organometallic Modelling of the Hydrodes-
ulphurization and Hydronitrogenation Reactions; Kluwer Academic
Publishers: Dordrecht, 2002.
(
14) Churchill, D. G.; Bridgewater, B. M.; Parkin, G. J. Am. Chem. Soc.
2
000, 122, 178.
(18) Garc ´ı a, J.; Maitlis, P. M. J. Am. Chem. Soc. 1993, 115, 12200. Garc ´ı a,
J.; Mann, B. E.; Adams, H.; Bailey, N. A.; Maitlis, P. M. J. Am. Chem.
Soc. 1995, 117, 2179. (b) Iretskii, A.; Adams, H.; Garc ´ı a, J. J.; Maitlis,
P. M. J. Chem. Soc., Chem. Commun. 1998, 61. Ar e´ valo, A.; Bern e` s,
S.; Garc ´ı a, J. J.; Picazo, G.; Maitlis, P. M. Organometallics 1999, 18,
1680.
(
15) Jones, W. D.; Chin, R. M. Organometallics 1992, 11, 2698.
16) Myers, A. W.; Jones, W. D. Organometallics 1996, 15, 2905. Paneque,
M.; Taboada, S.; Carmona, E. Organometallics 1996, 15, 2678.
17) Selnau, H. E.; Merola, J. S. Organometallics 1993, 12, 1583. Bianchini,
C.; Meli, A.; Peruzzini, M.; Vizza, F.; Frediani, P.; Herrera, V.;
S a´ nchez-Delgado, R. A. J. Am. Chem. Soc. 1993, 115, 2731. Jones,
W. D.; Chin, R. M. J. Am. Chem. Soc. 1994, 116, 198. (c) Bianchini,
C.; Jimenez, V.; Meli, A.; Vizza, F.; Herrera, V.; S a´ nchez-Delgado,
R. A. Organometallics 1995, 14, 2342. Paz-Sandoval, M. A.; Cer-
vantes-Vasquez, M.; Young, V. M., Jr.; Guzei, I. A.; Angelici, R. J.
Organometallics 2004, 23, 1274.
(
(
(19) (a) Vicic, D. A.; Jones, W. D. J. Am. Chem. Soc. 1997, 119, 10855.
(b) David, A. Vicic.; William, D. Jones. J. Am. Chem. Soc. 1999,
121, 7606. (c) Vicic, D. A.; Jones, W. D. Organometallics 1998, 17,
3411.
(20) Ate s¸ in, T. A.; Ate s¸ in, A. C.; Skugrud, K.; Jones, W. D. Inorg. Chem.
2008, 47, 4596.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010 12413

A R T I C L E S
Grochowski et al.
Scheme 1
with thiophene perpendicular to the plane. The Ni-P bond
lengths of 2.17 Å are slightly shorter than the 2.22 Å found for
2a.
Observation of Intermediates at Low Temperature. No
intermediates in the reaction of 1 with 2-cyanothiophene were
readily apparent at room temperature due to the rapidity of the
reaction. When 4 equiv of 2-cyanothiophene was added to a
2
Figure 1. Crystal structure of [(dippe)Ni(η -C,S-2-cyanothiophene)] (2a).
Ellipsoids are shown at the 50% probability level. Selected bond lengths
Å): Ni(1)-C(16), 1.9351(17); Ni(1)-S(1), 2.1590(5); Ni(1)-P(1), 2.2211(5);
Ni(1)-P(2), 2.2238(5); C(16)-C(17), 1.384(2); C(17)-C(18), 1.415(3);
C(18)-C(19), 1.349(3); C(19)-S(1), 1.702(2). Selected bond angles (°):
S(1)-Ni(1)-C(16),94.44(6);P(2)-Ni(1)-S(1),83.955(19);P(2)-Ni(1)-P(1),
(
8
solution of 1 at -60 °C in THF-d , however, several intermedi-
8
6.398(18); P(1)-Ni(1)-C(16), 95.29(6).
3
1
ates were observed while monitoring the reaction by P NMR
spectroscopy, as summarized in Scheme 1. 2a was present but
in a small proportion (9%) compared to the other species. A
direct relationship between the P-P coupling constant and the
nickel oxidation state has been noted in previous studies of
1
1b
reactions involving 1. In these studies, a Ni(0) complex is
typified by a P-P coupling constant of J > 60 Hz, while Ni(II)
complexes are typified by J < 40 Hz. In addition to the small
quantity of 2a, a second Ni(II) complex was present (19%) with
doublet resonances at δ 85.29 and 74.29 (JP-P ) 23 Hz). This
product appears to be the C-S-activated complex arising from
2
insertion into the nonsubstituted C-S bond, (dippe)Ni(κ -S,C-
SC(CN)dCHCHdCH) (2b). Theoretical calculations lend sup-
port to this assignment (vide infra). NMR integration showed
the ratio of 2b:2a at -60 °C to be 2:1. This observation parallels
the reaction involving the Pt-dippe analogue, in which activa-
tion of the nonsubstituted C-S bond of 2-cyanothiophene is
observed as a kinetic product, with eventual conversion to the
thermodynamically favored activation of the nitrile-substituted
Figure 2. Crystal structure of (dippe)Ni(2-thienyl)(CN) (3). Ellipsoids are
shown at the 50% probability level. Selected bond lengths (Å): Ni(1)-C(16),
1
.918(3); Ni(1)-C15, 1.877(2); Ni(1)-P(1), 2.1730(7); Ni(1)-P(2), 2.1821(7);
C(16)-C(17), 1.373(13); C(17)-C(18), 1.397(15); C(18)-C(19), 1.342(6);
C(19)-S(1), 1.737(5); S(1)-C(16), 1.729(7); C(15)-N(1), 1.137(3). Se-
lected bond angles (°): P(1)-Ni(1)-P(2), 88.49(3); P(1)-Ni(1)-C(16),
2
0
C-S bond.
9
2.3(4); C(16)-Ni(1)-C(15), 90.6(4); P(2)-Ni(1)-C(15), 88.48(7).
Two Ni(0) products were present, one displaying a pair of
doublets at δ 79.99 and 68.73 (JP-P ) 62 Hz) (4) and a second
displaying a set of doublets at δ 64.16 and 60.03 (JP-P ) 75
Hz) (5a). 4 was a major component of the mixture (29%) and
2
similar to those of (dippe)Ni(κ -C,S-C
4
H
4
S). 2a was isolated
as a solid that is stable under nitrogen. Exposure to air caused
the compound to decompose, and dippe oxide could be observed
by P NMR spectroscopy (δ 52.9, THF). When 2a was heated
in THF at 85 °C over a period of 13 h, transformation of the
compound occurred to give the C-CN bond-activated product
2
3
1
is proposed to be (dippe)Ni(η -C,N-2-cyanothiophene) (Scheme
1
). This complex was fairly stable and did not disappear until
the reaction temperature was raised to 0 °C (Figure 3).
Theoretical calculations (vide infra) and experimental precedent
(
dippe)Ni(CN)(2-thiophene) (3) (eq 6). Over the course of the
were used in the assignment of this structure. Previously, when
2
1
was reacted with 2-cyanopyridine, the η (C,N) complex
21
31
formed and was structurally characterized. The P NMR data
showed doublets at δ 82.3 and 68.6 (JP-P ) 68 Hz), very similar
to those found for 4. A previous study of the reaction with
2
benzonitrile showed formation of an η (C,N) complex that was
structurally characterized and which showed doublet resonances
3
1
22
at δ 78.6 and 66.4 (JP-P ) 68 Hz) in the P NMR spectrum.
reaction, some Ni(dippe)(CN)
2
precipitated out of solution as a
byproduct, and doublet resonances for 3 were observed at δ
5.64 and 76.89 (JP-P ) 31 Hz). The isolated product (74%)
1
0a
Both of these reactions also went on to give their respective
C-CN activated products. The 2-cyanothiophene reaction
8
was crystallized from benzene/hexanes, and X-ray diffraction
was used to obtain a crystal structure of 3 (Figure 2), showing
the expected square planar geometry around the nickel center,
(
21) Garcia, J. J.; Brunkan, N. M.; Jones, W. D. J. Am. Chem. Soc. 2002,
24, 9547.
(22) Garcia, J. J.; Jones, W. D. Organometallics 2000, 19, 5544.
1
1
2414 J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010

C-S vs C-C Bond Cleavage in 2-Cyanothiophene with [Ni(dippe)H]
2
A R T I C L E S
therefore has a reaction pathway similar to those of
1
1b
11
2
-cyanopyridine
and benzonitrile, although 4 is more
2
transient than the other previously studied η (C,N) species. 5a
was a short-lived Ni(0) intermediate that disappeared after 10
2
min at -60 °C; it was assigned to the η -C,C adduct.
Another major component of the mixture that was observed
in the NMR spectrum had broad resonances present at δ 69.27
and 61.46. As the reaction temperature was increased, the
resonances became narrower and splitting was then observed
(JP-P ) 17 Hz). This complex is assigned as the dinuclear
2 3
complex [(dippe)Ni]
2
(µ-κ :η -1-cyanobutadienylthiolate) (6b),
analogous to the known fluxional parent complex without a
1
9b
cyano group (eq 3). Further evidence for assignment of this
structure as being derived from the kinetic insertion product
2
b is described below.
b was quite stable, based upon its persistent presence in the
P NMR spectra (Scheme 1, Figure 3). 6b formed at -60 °C
and was unreactive until it was the only product present besides
6
3
1
Figure 3. Temperature dependence of the intermediates in Scheme 1.
2
a at 0 °C. Compounds 2a, 2b, 4, and 6b apparently form a
steady-state equilibrium at -60 °C, as their relative concentra-
tions rapidly stabilize. When the reaction was warmed to -20
°C, 2b was consumed while 2a grew in. There was a small
decrease in 4 and no change in 6b. The reaction was then
warmed to 0 °C, whereupon 2b was consumed immediately
while 4 began to be rapidly consumed as 2a grew in proportion-
ally. 6b did not change its percent composition after 95 min at
0
°C, and if only 1 equiv of 2-cyanothiophene is employed per
two Ni(dippe) fragments, 6b is stable at room temperature for
days. As the reaction was warmed to room temperature, 6b
reacted with excess 2-cyanothiophene to give 2a as the only
product.
Evidence for a Binuclear Intermediate 6b. The identity of
6
b was established primarily by independent synthesis. It was
first thought that 6b may have been 2b, given its small P-P
coupling constant (J ) 17 Hz). To explore this possibility, 1
was reacted with 2,5-dicyanothiophene, since cleavage of either
C-S bond would give the same product. When the reaction
was performed at room temperature, a long-lived intermediate
2
Figure 4. Crystal structure of (dippe)Ni(η -C,S-1,4-dicyanobutadienylthi-
olate) (2c). Ellipsoids are shown at the 50% probability level. Selected bond
lengths (Å): Ni(1)-C(16), 1.926(5); Ni(1)-S(1), 2.1444(16); Ni(1)-P(1),
2.2005(16);Ni(1)-P(2),2.2270(16).Selectedbondangles(°):S(1)-Ni(1)-C(16),
(
6c) was seen with broad resonances at δ 69.05 and 58.98 that
9
4.73(17); P(2)-Ni(1)-S(1), 84.60(6); P(2)-Ni(1)-P(1), 86.88(6);
were very similar to those observed for 6b (eq 7). The final
P(1)-Ni(1)-C(16), 93.74(17).
was also carried out by reacting 6c with a stoichiometric amount
of chloroform. The chloroform acted as a quenching agent and
gave 2c and Ni(dippe)Cl
2
. These products are consistent with
the proposed structure for 6c in which the coordinating
0
II
2
Ni (dippe) fragment is converted to Ni (dippe)Cl .
2
Independent synthesis of 6b proved to be more challenging
than that of 6c. Reaction of 2a with excess 1 (0.75 equiv) gave
product was the C-S insertion product (dippe)Ni(κ -S,C-1,4-
dicyanobutadienylthiolate)] (2c), with a virtual quartet centered
at δ 76.17 (JP-P ) 37 Hz). X-ray diffraction was employed to
confirm its structure, which is similar to that of 2a, with square
planar geometry around Ni and nearly identical bond lengths
and angles (Figure 4).
6
b, as seen in the low-temperature reaction, but also gave an
additional product (6a) with a pair of broad resonances in the
31
P NMR spectrum (eq 8). The two sets of peaks were proposed
1
The H NMR spectrum of 6c supports the proposed dinuclear
structure of the nickel(II) butadienethiolate metallacycle coor-
dinated to a Ni(0)-dippe fragment. The two butadienyl reso-
nances were observed at δ 4.94 and 7.80, the upfield shift being
1
3
indicative of π-coordination to a Ni(0) center. The C NMR
spectrum also demonstrated this phenomenon, with two of the
butadienethiolate resonances shifted upfield to δ 99.36 and
to correspond to isomeric dinuclear compounds, 6a and 6b, in
a 3:1 ratio. A corresponding set of H NMR resonances was
1
7
7.77, compared to the corresponding resonances in 2c, where
the farthest upfield vinylic carbon peaks occurred at δ 111. 6c
is a black solid formed in quantitative yield, and its composition
was confirmed by elemental analysis. Chemical characterization
seen for each product with the same 3:1 ratio (Figure 5).
H NMR resonances for the butadienylthiolate protons of 6b
were observed at δ 4.76, 5.68, and 7.51 in both the low-
1
J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010 12415

A R T I C L E S
Grochowski et al.
1
31
A broadband H{ P} decoupling experiment showed the
following splittings for 6b: H (doublet), H (triplet), and H
doublet). A COSY NMR spectrum showed that H was coupled
to H , and H was coupled to H and H , while H was coupled
to H . The agreement of the H{ P} experiments and COSY
experiments supports the proposed structure of 6b. A COSY
spectrum of 6a showed that the triplet resonance at δ 7.08 (H
was coupled to the peak at δ 4.25 (H , doublet) and the peak at
δ 5.35 (H , doublet). The splittings of H and H were invariant
with all attempted H{ P} decoupling experiments. A very
A
B
C
(
A
B
B
A
C
C
1
31
B
D
)
E
F
D
F
1
31
E
slight change was seen for H , which was tentatively assigned
to the resonance at δ 4.25.
A dramatic difference between 6b and 6a is the shape of the
resonances in the 31P NMR spectrum. 6a showed resonances
that were much broader than those of 6b, although the chemical
shifts were similar. Fluxional exchange is slower in 6a as
compared to 6b, which accounts for the broadness of the
resonances. When a solution containing 6a was cooled to -40
°
C, the two resonances were observed as four broad resonances,
consistent with the freezing out of the fluxional process (as seen
1
9b
in the parent complex with thiophene ) and indicative of the
low symmetry of 6a (eq 9). The position of the cyano group
1
Figure 5. (Top) H NMR spectrum of thienyl peaks for eq 8. (Bottom)
31
P NMR of dippe peaks for eq 8. THF-d , 25 °C.
8
for each isomer may account for the exchange rate differences
between 6b and 6a; i.e., 6a, with a cyano substituent on the
double bond π-bound to Ni(0), is less labile and its static
structure can be frozen out. In addition, the cyano group
stabilizes 2b over 6b plus 2-cyanothiophene, in contrast to the
parent thiophene insertion complex, which establishes an
equilibrium between the mononuclear complex and dinuclear
Figure 6. (Left) 6b-H
A A
before decoupling. (Middle) Irradiation of 6b-P
nuclei at δ 68.38 results in resolution to a triplet of doublets. (Right)
1
31
A
Broadband H{ P} resolves 6b-H to a doublet.
temperature reaction of 1 with 2-cyanothiophene and the
independent synthesis described above, where it appeared to
be the minor product. Similar resonances were observed for 6a
at δ 4.25, 5.35, and 7.08. Identification and differentiation of
1
9b
complex plus thiophene.
Computational Studies: Ground States and Transition
6
b and 6a were made on the basis of the splitting patterns of
the resonances. For 6b, complex patterns were observed at δ
.76 (H ) and especially δ 5.68 (H ), while splitting patterns
States. As described above, the low-temperature reaction of
[(dippe)NiH] with 2-cyanothiophene formed five different
2
4
B
A
complexes, including two C-S activation products. When the
same reaction was conducted at 85 °C, only the C-CN cleavage
product was observed. On the basis of the experimental results,
a computational study was undertaken to examine the energetics
and reaction pathways leading to both C-S and C-CN
oxidative addition products. To simplify the calculation, the
[Ni(dmpe)] fragment was used as a model of the active species
[Ni(dippe)], with methyl groups replacing the isopropyl
for 6a were simpler. The proposed structure for 6b would allow
for coupling between phosphorus nuclei and the thienyl protons,
while such coupling would be largely absent for 6a due to a
larger through-bond distance between the nuclei. Additional
support for the assignment of 6a and 6b came from reaction of
the mixture with chloroform at room temperature, which led to
the formation of 2b and Ni(dippe)Cl
c.
A series of H{ P} decoupling experiments supports the
2
, similar to the reaction of
2
3
6
groups. Initially, Ni(dmpe)(2-thienyl)(CN) (S3) and Ni(d-
1
31
2
ippe)[κ -C,S-SCHCHCHC(CN)] (S2a) were constructed on the
3
1
assigned structures. Irradiation of the P resonance at δ 68.38
6b-P ) resulted in resolution of the peak at 5.68 (6b-H ) into
a triplet of doublets (Figure 6). Splitting by the two equivalent
basis of the X-ray crystallographic structures of 3 and 2a
(Figures 1 and 2), and local minima were found in gas phase
calculations.
Both of the optimized structures of S3 and S2a have nearly
square planar geometries around nickel coordination sites. The
Ni-CN and Ni-C(thiophene) distances in S3 are 1.868 and
1.915 Å, respectively, similar to those in the crystal structure
of 3 (1.877 and 1.918 Å). In S2a, the Ni-S bond is 2.195 Å,
only 0.036 Å longer than that in the crystal structure of 2a.
(
A
A
3
P
A
nuclei gave a triplet ( JP-H ) 18 Hz), which was further
split into a doublet (JH-H ) 8 Hz) by the adjacent vicinal proton
3
1
at δ 4.76 (6b-H
b (P ) at δ 60.99 resulted in 6b-H
rounded peak, implying that the set of P
residual splitting to H . H resolved to a triplet, implying
residual coupling with P nuclei splitting by each of the two
adjacent thienyl protons. The doublet at 7.51 (6b-H ) was not
affected by H{ P} decoupling and was observed as a doublet.
B
). Irradiation of the second P resonance of
being observed as a
nuclei gives some
6
B
A
B
A
B
B
C
(
23) Atesin, T. A.; Li, T.; Jones, W. D. J. Am. Soc. Chem. 2007, 129,
1
31
7562.
1
2416 J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010

C-S vs C-C Bond Cleavage in 2-Cyanothiophene with [Ni(dippe)H]
2
A R T I C L E S
Table 1. Optimized Structures (Interatomic Distances in Å, Angles in Deg) and Relative Free Energies (∆G Relative to Fragments, kcal/mol,
a
PCM-Corrected in THF) of Stationary Points on the [Ni(dmpe)] + 2-cyanothiophene Potential Energy Surface
Ni-S
Ni-C1
Ni-C2
Ni-C3
Ni-C4
Ni-CN
Ni-N
C-CN
C-N
C1-S
C4-S
R
ꢀ
∆G
S2a
S2b
S3
S4
S5a
S5b
S7a
S7b
TS2a5a
TS2b7b
TS35a
TS45a
TS5a5b
TS5b7b
2.195
2.174
1.929
1.424
1.430
1.172
1.166
1.171
1.238
1.169
1.167
1.169
1.171
1.169
1.169
1.19
1.729
13.8
3.7
78.9
1.4
18.9
-50.36
-44.06
-44.3
-38.88
-33.57
-32.04
-18.7
-25.09
-19.84
-22.47
-12.09
-7.15
-20.71
-24.15
1.890
1.751
1.755
1.763
1.822
1.783
1.908
1.771
2.001
1.773
1.78
1.915
1.956
1.998
1.959
1.868
1.872
1.748
1.735
1.766
1.810
1.743
1.807
1.738
2.179
1.739
1.746
1.787
1.802
1.4
1.6
1.898
1.441
1.425
1.414
1.428
1.401
1.426
1.413
1.561
1.412
1.412
1.402
1.977
66.1
73.0
60.2
76.4
64.6
64.8
74.6
88.3
87.7
83.8
2.000
1.946
1.960
1.907
2.231
2.280
2.214
2.187
1.967
2.427
2.293
1.883
2.003
34.9
62.8
2.366
1.188
1.169
1.171
1.766
1.813
1.777
2.046
2.090
2.562
2.309
1.928
2.346
a
R is the angle between P-Ni-P and thiophene planes. ꢀ is the angle between P-Ni-P and Ni-C-CN planes. The numbering of carbon atoms on
the thiophene ring is shown in the thiophene fragment (first structure) in Figure 7 or Figure 11.
With the simplified model, the DFT optimized structures are
still consistent with the crystallographic data for the analogous
dippe compounds.
The other possible C-S activation product in which cleavage
occurs away from the cyano substituent was not stable enough
to be isolated. Consequently, the model complex S2b was
constructed starting from the crystal structure of 2a and
optimized also as a ground state. The transition states leading
to the three oxidative addition products S3, S2a, and S2b were
located from relaxed potential energy surface scans by con-
straining the C-S or C-CN distance in the products and
optimizing the resulting structures during the processes. Those
structures around the energy maxima were optimized as
transition states, and the intrinsic reaction coordinate (IRC)
routes were calculated in both forward and reverse directions
toward the corresponding local minima. The transition state
TS35a connects the C-CN cleavage product S3 to a complex
2
S5a with nickel η -coordinated through the CdC of the
thiophene ring next to the CN substituent. The carbon-carbon
bond distance is lengthened to 1.458 Å due to back-donation
from the nickel center. In the same way, TS2a5a was found to
connect one of the C-S activation products S2a also to S5a.
TS2b7b, on the other hand, connects the other C-S activation
product S2b to a complex S7b, in which the nickel metal
coordinates to the C-S bond (a shallow minimum). The C-S
distance is elongated from 1.807 Å in S7b to 2.179 Å in
TS2b7b. When further constraining the C-S distance in S7b,
another transition state, TS5b7b, was found. It connects S7b
2
to another local minimum, S5b, that has nickel η -coordinated
to the CdC of the thiophene ring away from the CN substitute,
similar to S5a. Complex S7a, the counterpart of S7b in the other
C-S activation, is slightly higher in energy than TS2a5a and
therefore was excluded from the C-S activation pathway
leading to S2a. The nickel metal can also migrate from S5a to
4
S5b via an η -coordinated transition state, TS5a5b, which makes
possible the intramolecular interchange between the two C-S
activation products. All of the optimizations and potential energy
surface scans were conducted in the gas phase. Solvent effects
Figure 7. Optimized structures (interatomic distances in Å, angles in deg)
on the relative stabilities of all structures were taken into
consideration by applying a polarizable continuum model (PCM)
correction. The optimized structural parameters and relative
energies in THF of all ground states and transition states found
in this study are summarized in Table 1 and Figure 7 (see movies
for IRC paths in Supporting Information).
of stationary points and transition states on the [Ni(dmpe)] + 2-cyan-
othiophene potential energy surface. Hydrogen atoms and methyl groups
on the phosphorus have been omitted.
Kinetic and Thermodynamic C-S Activations. From the
experimental observations, the C-S activation product 2a is
thermodynamically more stable than 2b. From DFT calculations
J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010 12417

A R T I C L E S
Grochowski et al.
Figure 8. Total free energies (PCM-corrected in THF).
(
PCM corrected in THF) as shown in Figure 8, S2a lies 6.3
by 0.45 Å in TS2b7b. Consequently, the C-S bond is seen to
kcal/mol lower in energy than S2b. When changing the
Ni(dmpe) to the Ni(dippe) fragment, DFT calculations showed
that 2a is still about 3 kcal/mol downhill from 2b. In agreement,
at room temperature, no resonances for 2b could be seen in the
be well along the way to being broken, especially in the kinetic
product 2b. The C-S bond is also close to lying in the NiP
2
plane (14° in TS2a5a and 5° in TS2b7b). These transition states
stand in contrast to the transition state for C-CN bond cleavage.
In TS35a, the Ni-CN and Ni-C bonds are also completely
formed (within 0.05 Å of their values in 3). The C-CN bond
is stretched by only 0.14 Å, so the bond cleavage is not nearly
as well progressed as in the case of the C-S cleavage. Also,
2
NMR spectrum. On the other hand, from the η -C,C-2-
cyanothiophene complexes, the activation barrier for the C-S
activation leading to S2a is 13.7 kcal/mol, and that leading to
S2b is only 9.6 kcal/mol. Therefore, the calculations are in
agreement with the experimental results that S2b is the kinetic
product while S2a is the thermodynamic product. This also
matches the previous results obtained with Pt(dmpe) and
2
the C-CN bond lies at an angle of 31° to the NiP plane, similar
1
1d
to the value seen in the cleavage of benzonitrile (28°), and
indicates that the reaction is less progressed toward the square
planar product than in the case of the C-S cleavage.
2
0
2
-cyanothiophene, in which a kinetic preference for cleavage
of the unsubstituted C-S bond was observed.
One can also see why there is a kinetic preference for C-S
cleavage over C-CN cleavage by examination of the HOMOs
and LUMOs for the [Ni(dmpe)] fragment and the 2-cyan-
othiophene. As shown in Figure 9, the HOMO of the nickel is
largely dx2-y2, and donation into the 2-cyanothiophene LUMO
C-S vs C-CN Activations. Upon heating, the C-S activation
product 2a converted completely to the C-CN oxidative
addition product 3. When the system was cooled back to room
temperature, 2a was not formed again, indicating that 3 is
thermodynamically more stable than 2a. However, the energetic
results from DFT calculations for these two complexes contra-
dict this experimental observation. S3 was found to lie 6.8 kcal/
mol higher in energy than S2a in the gas phase. Even after using
PCM to correct the solvent effects, S3 was still higher in energy
than S2a by 8 kcal/mol in benzene and 6.1 kcal/mol in THF
would lead to population of an orbital that is C-S antibonding,
thereby promoting C-S cleavage. The interaction of the
2-cyanothiophene HOMO with the [Ni(dmpe)] LUMO results
2
in formation of complexes S5a and S5b, the η -C,C adducts of
the thiophene that are the immediate precursors to the transition
states for C-S cleavage. Both of these interactions are consistent
with a kinetic preference for C-S cleavage.
(Figure 8). At first, the simplified model Ni(dmpe) was thought
to cause the error in the relative energy. Therefore, the (dippe)Ni
analogues 3 and 2a were also optimized with DFT calculations.
The energy difference between 3 and 2a decreased to 3.4 kcal/
mol in the gas phase, 3.6 kcal/mol in benzene, and only 1.4
kcal/mol in THF. The C-CN activation product 3 was still
uphill in energy compared to the C-S activation product 2a.
The reason for this inconsistency is not clear, but the trend
indicates that 3 is approaching the same energy as 2a.
It is worthwhile to examine the calculated transition states
for C-S and C-C cleavage. The two transition states for C-S
cleavage, TS2a5a and TS2b7b, show completely formed Ni-C
and Ni-S bonds, within 0.03 Å of their final values in S2a and
S2b. The C-S bond is lengthened by 0.25 Å in TS2a5a and
Dinuclear Complexes. In the presence of excess [Ni(d-
ippe)H]
peaks in both the H and P NMR spectra, with characteristics
similar to those of the dinuclear complex [(dmpe) Ni
However, crystal
structures of these products could not be obtained. Therefore,
DFT models of the [(dmpe) Ni (2-cyanothiophene)] complexes
were constructed on the basis of the crystallographic structure
of [(dmpe) Ni (benzothiophene)] (Figure 10) by changing the
2
, the reaction at low temperature gave two new sets of
1
31
2
2
-
1
9b
(benzothiophene)] reported previously.
2
2
2
2
bridging ligand from benzothiophene to 2-cyanothiophene. Due
to the asymmetric molecular structure of the 2-cyanothiophene
substrate, two possible dinuclear complexes have been con-
structed and optimized as ground states (S6a and S6b).
1
2418 J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010

C-S vs C-C Bond Cleavage in 2-Cyanothiophene with [Ni(dippe)H]
2
A R T I C L E S
Figure 11. Optimized structures (interatomic distances in Å, angles in deg)
of dinuclear complexes [(dmpe) Ni (2-cyanothiophene)] (S6a, S6b) and their
2 2
dippe analogues (6a, 6b). Hydrogen atoms and isopropyl groups on the
phosphorus have been omitted.
Figure 9. HOMO-LUMO interactions between [Ni(dmpe)] and 2-cyan-
othiophene. Energies are in eV.
for S6a and S6b, as well as 6a and 6b. So far no evidence has
been found to prove that the actual geometries are the same as
in those in Figure 11. Three other dinuclear complexes have
also been located, but all of them have much higher energies
than S6a and S6b (see Supporting Information).
Conclusion
C-S bond activation occurred cleanly at room temperature
when [Ni(dippe)H] (1) was reacted with 2-cyanothiophene,
2
producing a Ni metallacycle product (2a) where bond cleavage
occurs on the nitrile-substituted side of the thiophene. When
2
a was heated in solution at 85 °C, conversion to the C-CN
activated product 3 occurred, demonstrating that C-S cleavage
is kinetically favored over C-C cleavage, whereas the latter is
thermodynamically preferred. When the reaction was monitored
at -60 °C by NMR spectroscopy, several intermediates were
observed to be in equilibrium, including the isomer of 2a in
2 2
Figure 10. X-ray structure of (dippe) Ni (benzothiophene). (Reproduced
with permission from ref 19b.)
2
which the alternate C-S bond had been cleaved (2b), an η -
Structures and selected bond lengths are shown in Figure 11
and Table 2. S6b is formed with Ni(dmpe) and the kinetic C-S
cleavage product S2b, while S6a is formed with Ni(dmpe) and
the thermodynamic C-S activation product S2a. Consistently,
S6a is thermodynamically more stable than S6b. From the
optimized structures, in each dimer there are four inequivalent
phosphorus environments. For 6a, four distinct resonances
appeared in the 31P NMR spectrum at low temperature (-40
C,N complex of 2-cyanothiophene (4), and a dinuclear adduct
of 2b (6b). Activation from the nonsubstituted side of the
thiophene was predominant at low temperature over 2a at -60
°C (i.e., 2b is kinetically preferred), indicating steric contribu-
tions to the kinetic selectivity. This selectivity is similar to the
previously reported reaction of 2-cyanothiophene with the
analogous Pt-dippe system, which showed that cleavage of the
nonsubstituted C-S bond initially predominates, but over time
conversion to cleavage of the substituted C-S bond occurs as
the major thermodynamic product. The Ni system mirrors this
pathway (and energetics), except that the reaction is much faster
°C). For 6b, however, only one pair of doublets was observed,
suggesting that there is a rapid exchange between the two nickel
coordination sites that makes the two dippe ligands
1
9b
2
equivalent.
movie of exchange in Supporting Information).
When [Ni(dippe)H] was reacted with 2a at room temperature,
This interchange has also been modeled (see
than that of the Pt system. An η (C,N) intermediate (4) was
also observed, comparable to that seen in analogous reactions
of 1 with 2-cyanopyridine and benzonitrile, yet this species is
not critical to the C-S or C-C cleavage reactions. Another
intermediate seen in the low-temperature reaction was the
dinuclear complex 6b. This mixed-valent Ni(II)-Ni(0) complex
was found by both experiment and theory to be more thermo-
dynamically stable than 2a plus [Ni(dippe)].
Independent synthesis of 6b also led to the discovery of the
regioisomer 6a, which was not observed in the reaction of
2
[Ni(dippe)H] with excess 2-cyanothiophene. DFT calculations
2
both 6a and 6b were observed and formed an equilibrium with
a ratio of 3:1. Thus, the Gibbs free energy of 6b should be about
0
.6 kcal/mol higher than that of 6a. However, using the
Ni(dmpe) model, DFT calculations showed that the energy of
S6b in THF was around 7 kcal/mol higher than that of S6a.
Even after optimizations with the complete Ni(dippe) fragment,
the energy difference between dinuclear complexes 6a and 6b
was still 6 kcal/mol in the gas phase and 4 kcal/mol in THF.
The error in energy might come from use of the model structures
were used to model this reactivity and indicated that the greater
J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010 12419

A R T I C L E S
Grochowski et al.
Table 2. Optimized Structures (Interatomic Distances in Å, Angles in Deg) and Relative Free Energies (∆G Relative to Free Fragments,
kcal/mol, PCM-Corrected in THF) of Dinuclear Complexes on the 2[Ni(dmpe)] + 2-Cyanothiophene Potential Energy Surface
Ni1-S
Ni2-S
Ni1-Ni2
Ni1-C1
Ni1-C4
Ni2-C1
Ni2-C2
Ni2-C3
Ni2-C4
∆G
S6b
S6a
2.264
2.282
2.276
2.263
2.462
2.728
2.354
3.347
2.650
2.854
2.736
3.403
1.902
2.254
1.968
-73.76
-80.78
-64.27
-68.19
1.944
1.974
1.958
1.995
2.056
2.005
6
6
b
a
1.921
2.463
1.990
(s), 19.23 (s), 18.27 (s). 31P{ H} NMR (162 MHz, THF-d
δ 85.64 (d, J ) 31 Hz), 76.89 (d, J ) 31 Hz). Anal. Calcd (found)
S: C, 53.05 (52.93); H, 8.20 (7.88); N, 3.26 (3.13).
Low-Temperature Reaction of [Ni(dippe)H] with Excess
-Cyanothiophene. In a drybox, 1 (9.4 mg, 0.0146 mmol) was
dissolved in 1 mL of THF-d and placed in an NMR tube. A septum
1
ease of C-S activation over C-C activation can be attributed
to a later transition state for cleavage in the latter, in which
greater metal-carbon bonding is present.
8
, 25 °C):
for C19H35NNiP
2
2
2
Experimental Section
8
General Procedures. All operations were performed under a
was placed on the tube and wrapped with parafilm. The NMR tube
was placed in the NMR probe and cooled to -60 °C. 2-Cyan-
othiophene (5.4 µL, 0.058 mmol) was dissolved in 0.1 mL of THF-
d8, the solution was drawn into a syringe, and the needle was capped
with a septum to minimize the exposure to air. The solution was
injected into the cold NMR tube containing 1 and immediately
placed back in the probe. An initial P NMR spectrum was taken,
and subsequent spectra were taken over the course of 65 min. A
steady-state equilibrium was established at this temperature between
2a, 2b, 4, and 6b. The probe was then heated to -20 °C and
monitored by NMR spectroscopy over the course of 40 min, and
then heated to 0 °C and monitored by NMR spectroscopy over the
course of 90 min. Observed changes are described in the text, and
data are given below.
nitrogen atmosphere unless otherwise stated. The complex [Ni(d-
1
9a
2
ippe)H] was prepared as previously reported. 2-Thiophenecar-
bonitrile (99%) was purchased from Aldrich Chemical Co., dried
by stirring with sodium sulfate, and then stored under nitrogen.
24
2,5-Thiophenedicarbonitrile was prepared as previously reported.
3
1
THF-d and benzene-d were distilled from sodium/benzophenone
8
6
prior to use. A Bruker-AXS SMART platform diffractometer
equipped with an APEX II CCD detector was used for X-ray crystal
structure determination. Elemental analyses were obtained from
1
31
13
Desert Analytics. All H, P, and C spectra were recorded on
1
Bruker Avance 400 and 500 MHz spectrometers; all H chemical
shifts are reported relative to the residual proton resonance in the
deuterated solvent, and 31P chemical shifts are referenced relative
2
to an 85% H
3
PO
4
external standard.
[(dippe)Ni(K -S,C-SC(CN))CHCHdCH] (2b). 2b was ob-
2
31
Preparation of [(dippe)Ni(K -S,C-SCHdCHCHdC(CN)]
served at -60 °C by P NMR spectroscopy, where it was present
(
2a). In a nitrogen-filled drybox, 2-cyanothiophene (5.7 µL, 0.061
in a 2:1 ratio over 2a. When the reaction temperature was raised
to -20 °C, 2b started to decrease in the reaction mixture. By the
time the probe temperature had been raised to 0 °C, 2b was no
mmol) was added by syringe to a J-Young NMR tube containing
(15.8 mg, 0.0245 mmol) in 1.5 mL of THF. The blood-red
1
31
1
solution turned dark brown upon addition. After 12 h the solution
changed to red. The THF was removed by vacuum, and the reddish-
brown solid was dried by vacuum for at least 30 h to ensure removal
longer present. P{ H} NMR (162 MHz, THF-d , -60 °C): δ 85.29
8
(d, J ) 23 Hz), 74.91 (d, J ) 23 Hz).
2
[(dippe)Ni(η -C,N-2-cyanothiophene)] (4). 4 was observed at
1
31
of excess 2-cyanothiophene (19.7 mg, 0.0458 mmol, 93%). H NMR
-60 °C by P NMR spectroscopy as a significant component in
(
400 MHz, C
6
D
6
, 25 °C): δ 7.84 (dd, 1 H, J ) 14, 7 Hz), 7.43 (t,
the mixture. When the probe temperature was raised to -20 °C, 4
began to be consumed slowly. Consumption was more rapid when
the temperature was raised to 0 °C. After 95 min at 0 °C only a
1
7
1
H, J ) 10, 9 Hz), 6.88 (t, 1 H, J ) 8 Hz), 3.12 (sextet, 2 H, J )
Hz), 2.05 (sextet, 2 H, J ) 7 Hz), 1.28 (dd, 6 H, J ) 17, 7 Hz),
3
1
1
.21-0.89 (m, 4 H), 1.10 (dd, 6 H, J ) 15, 7 Hz), 0.86 (m, 12 H).
small amount of 4 remained. P{ H} NMR (162 MHz, THF-d
-60 °C): δ 79.99 (d, J ) 62 Hz), 68.73 (d, J ) 62 Hz).
8
,
1
3
1
C{ H} NMR (126 MHz, C
6
D
6
, 25 °C): δ 142.46 (s), 132.64 (dd,
2
3
J ) 17, 4 Hz), 128.90 (d, J ) 12 Hz), 120.62 (s), 113.47 (dd, J )
2
[(dippe)Ni] (µ-K :η -1-cyanobutadienylthiolate)] (6b). 6b was
1 31
7
6
3
2, 17 Hz), 24.39 (s), 24.22 (s), 24.13 (s), 23.98 (s), 21.70 (d, J )
observed at -60 °C by H and P NMR spectroscopy. It was not
consumed as the probe temperature was raised to 0 °C. At room
Hz), 20.49 (d, J ) 20 Hz), 20.30 (d, J ) 20 Hz), 19.56 (d, J )
Hz), 19.21 (d, J ) 22 Hz), 19.10 (d, J ) 21 Hz), 18.25 (s), 18.20
temperature it was consumed completely to give 2a as the only
1
1
(
s), 18.16 (s). 31P{ H} NMR (162 MHz, C
6
D
6
, 25 °C): δ73.06 (d,
product. H NMR (400 MHz, THF-d
8
, -20 °C): δ 7.51 (d, J ) 7
), 4.76 (m, JH-H
B
). The coupling constants were assigned by
J ) 35 Hz), 72.22 (d, J ) 35 Hz). Anal. Calcd (found) for
S: C, 53.05 (53.30); H, 8.20 (7.49); N, 3.26 (3.05).
Preparation of [(dippe)Ni(CN)(2-thiophene)] (3). In a nitrogen-
filled glovebox, 2a (20.5 mg, 0.0477 mmol) was dissolved in 1.5
mL of THF-d and placed in a J-Young NMR tube. The tube was
Hz, H
C
), 5.68 (m, JH-H ) 8 Hz, JP-H ) 18 Hz, H
A
19
C H
35NNiP
2
) 7,8 Hz, H
1
31
31
1
H{ P}experiments. P{ H}NMR (162 MHz, THF-d
δ 69.27 (d, J ) 17 Hz), 61.46 (d, J ) 17 Hz).
8
, -20 °C):
2
8
(dippe)Ni(η -C,C-2-cyanothiophene) (5a). 5a was a very minor
complex which was observed at -60 °C at the start of the low-
temperature reaction. It was assigned as a Ni(0) complex because
of its large P-P coupling constant (J ) 75 Hz). It was short-lived
31
heated in an 85 °C oil bath and monitored periodically by P NMR
spectroscopy. The reaction was completed after 13 h, at which point
the solvent was removed by vacuum. The residue was extracted
with benzene and filtered through a plug of Celite to remove the
3
1
1
and disappeared after 10 min at -60 °C. P{ H}NMR (162 MHz,
byproduct [(dippe)Ni(CN)
to afford a yellow solid (15.2 mg, 0.0353 mmol, 74%). H NMR
400 MHz, C
2
]. The benzene was removed by vacuum
THF-d
8
, -60 °C): δ 64.16 (d, J ) 75 Hz), 60.03 (d, J ) 75 Hz).
1
2
3
Reaction Producing Isomeric Forms of [(dippe)Ni] (µ-K :η -
2
(
6
D
6
, 25 °C): δ 7.62 (m, 1 H), 7.27 (s, 1 H), 7.03 (s,
H), 2.24 (sextet, 2 H, J ) 7 Hz), 1.92 (sextet, 2 H, J ) 7 Hz),
C,S-Cyanobutadienylthiolate) (6a,b)). 2a (24.0 mg, 0.0558 mmol)
was dissolved in 1 mL of THF-d and added to a weighed sample
of 1 (26.3 mg, 0.0408 mmol). The dark reddish-brown solution
1
1
8
13
1
.48-1.20 (m, 10 H), 0.98-0.81 (m, 18 H). C{ H} NMR (126
, 25 °C): δ 147.64 (dd, J ) 63, 26 Hz), 131.66 (s),
28.84 (s), 127.34 (s), 125.00 (d, J ) 78 Hz), 26.73 (d, J ) 23
Hz), 26.56 (s), 22.74 (d, J ) 19 Hz), 22.56 (d, J ) 19 Hz), 21.10
d, J ) 16 Hz), 20.93 (d, J ) 16 Hz), 20.52 (s), 20.15 (s), 19.71
MHz, THF-d
8
was transferred to a J-Young NMR tube. The reaction was complete
3
1
1
1
after 0.5 h, as confirmed by P and H NMR spectroscopies. The
ratio of 6a:6b was approximately 3:1 at room temperature. Data
2
3
(
for 6b are reported above. For [(dippe)Ni] (µ-κ :η -C,S-4-cyanob-
1
2
utadienylthiolate)] (6a), H NMR (400 MHz, THF-d
8
, 22 °C): δ
), 4.25 (d, J ) 5 Hz,
, 25 °C): δ 69.35 (br s), 59.25
7
H
.08 (t, J ) 6 Hz, H
E
D
), 5.35 (d, J ) 7 Hz, H
). P{ H}NMR (162 MHz, THF-d
F
(
24) Suzuki, H.; Iwao, T.; Sugiyama, T. Bull. Inst. Chem. Res., Kyoto UniV.
31
1
1
974, 52, 561.
8
1
2420 J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010

C-S vs C-C Bond Cleavage in 2-Cyanothiophene with [Ni(dippe)H]
2
A R T I C L E S
2
6
(
br s). When the probe temperature was lowered to -40 °C for 25
porating Becke’s three-parameter hybrid functional (B3), along
31
1
27
min, four broad peaks were then observed. P{ H}NMR (162 MHz,
THF-d , -40 °C): δ 77.50 (br s), 64.50 (br s), 59.50 (br s), 51.8
(
with the Lee-Yang-Parr correlation functional (LYP). All
2
8
8
calculations were performed using the Gaussian03 package. The
Ni and P atoms were represented with the effective core pseudo-
potentials of the Stuttgart group and the associated basis sets
13
br s). C NMR data for 6b and 6a: The alkyl region of the
spectrum is indecipherable due to the multiple overlapping reso-
2
29
nances and the presence of 1, 6b, and 6a in the sample. The sp
improved with a set of f-polarization functions for Ni (R ) 3.130)
resonances are distinct. 13C{ H} NMR (126 MHz, THF-d
1
8
, 25 °C):
and a set of d-polarization functions for P (R ) 0.387) and S (R )
30
δ 147.21 (s), 144.92 (s), 133.24 (s), 128.70 (s), 122.60 (s), 118.35
0.503). The remaining atoms (C, H, and N) were represented with
31
(
t, J ) 7 Hz), 97.80 (s), 96.56 (s), 83.55 (s), 80.71 (m).
6-31G(d,p) basis sets. The geometry optimizations were performed
without any symmetry constraints, and the local minima and the
transition states were checked by frequency calculations. For each
transition-state structure, the intrinsic reaction coordinate (IRC)
routes were calculated in both directions toward the corresponding
minima. Movies showing these motions are included in the
Supporting Information. For some of the transition states, the IRC
calculations failed to reach the energy minima on the potential
energy surface; therefore, in those cases geometry optimizations
were carried out as a continuation of the IRC path. Because of the
polarity of the structures, the solvent effects on their relative
stabilities were evaluated by calculating the free energies of
2
Preparation of (dippe)Ni(K -C,S-2,5-dicyanothiophene) (2c).
In a nitrogen-filled glovebox, 2,5-thiophenedicarbonitrile (6.6 mg,
.049 mmol) was dissolved in 1.5 mL of THF-d and added to 1
14.5 mg, 0.0225 mmol). The solution was placed in a J-Young
0
8
(
NMR tube and shaken to ensure proper mixing. The reaction was
monitored by 31P and H NMR spectroscopies and judged to be
complete after 1 day. The THF was evaporated under vacuum, and
the solid was dissolved in benzene and eluted through a plug of
Celite. The solution was concentrated to approximately 1 mL,
layered with pentane, and placed in a -30 °C freezer overnight.
1
The liquid was removed by pipet to leave a red-orange solid (12.1
1
9
mg, 0.0266 mmol, 59%). H NMR (400 MHz, THF-d
8
, 25 °C): δ
solvation in terms of the polarizable continuum model (PCM).
Self-consistent reaction field (SCRF) calculations using the PCM-
7
.26 (m, 1 H), 7.13 (d, 1 H, J ) 7 Hz), 3.04 (sextet, 2 H, J ) 7
32
Hz), 2.44 (sextet, 2 H, J ) 7 Hz), 2.03-1.84 (m, 4 H), 1.38-1.12
(
(
UA0 solvation model were carried out for the gas-phase optimized
structures. The dielectric constant in the PCM calculations was set
to ꢀ ) 7.58 to simulate THF and ꢀ ) 2.247 to simulate benzene as
the solvent media used in the experimental study. The energies
discussed in the potential energy surface plot are Gibbs free
energies, which have been calculated at 298.15 K and 1 atm. The
ORTEP32 package was used to display the molecular structures.
m, 24 H). 13C{ H} NMR (126 MHz, THF-d
1
, 25 °C): δ 142.45
8
s), 130.88 (s), 121.87 (dd, J ) 69, 20 Hz), 120.38 (d, J ) 5 Hz),
1
2
1
11.06 (s), 110.94 (s), 21.88 (d, J ) 5 Hz), 21.49 (d, J ) 26 Hz),
1.32 (d, J ) 26 Hz), 20.05 (d, J ) 2 Hz), 19.78 (d, J ) 23 Hz),
3
1
1
9.65 (d, J ) 24 Hz), 18.71 (s), 18.58 (d, J ) 5 Hz). P{ H}
NMR (162 MHz, THF-d
8
, 25 °C): δ 76.76 (d, J ) 37 Hz), 75.58
d, J ) 37 Hz). Anal. Calcd (found) for C20 NiP S: C, 52.77
52.45); H, 7.53 (6.83); N, 6.15 (5.83).
(
(
H
34
N
2
2
Acknowledgment. Financial support from the National Science
Foundation (CHE-0717040) for the experimental work (M.R.G.)
and from the U.S. Department of Energy (grant FG02-86ER13569)
for the calculational work (T.L.) is gratefully acknowledged.
2
Preparation of [(dippe)
6c). In a nitrogen-filled glovebox, 2,5-thiophenedicarbonitrile (5.0
and added
2 2
Ni (η -C,S-2,5-dicyanothiophene)]
(
mg, 0.038 mmol) was dissolved in 1.5 mL of THF-d
8
to 1 (24.2 mg, 0.0376 mmol). The black solution was placed in a
J-Young NMR tube and shaken to ensure proper mixing. After
Supporting Information Available: Table of distribution of
species for the low-temperature reaction; complete ref 28; figure
showing energies of calculated species in gas phase and benzene;
pictures of molecular orbitals; calculated coordinates; results
with additional basis sets; X-ray diffraction details (in CIF
format) for 2a, 3 (two morphologies), and 2c; and five movies
showing the IRC calculated reaction paths. This material is
available free of charge via the Internet at http://pubs.acs.org.
The structures are also deposited in the Cambridge Crystal-
lographic Database as CCDC Nos. 774682-774685.
1
2
.5 h, the reaction was judged to be complete by 31P and H NMR
spectroscopies. The THF was removed by vacuum, and the
crystalline black solid was dried for several hours (29 mg, 0.037
1
mmol, 99%). H NMR (400 MHz, THF-d
8
, 25 °C): δ 7.81 (d, 1 H,
J ) 7 Hz), 4.93 (br s), 2.78 (m, 2 H), 2.44 (m, 2 H), 2.18 (m, 4 H),
13
1
1
.51-0.94 (br m, 54 H). C{ H} NMR (126 MHz, THF-d
C): δ 157.05 (s), 148.52 (s), 135.91 (s), 122.28 (s), 99.36 (s), 77.77
s), 26.56 (s), 24.79 (d, J ) 13 Hz), 22.10 (d, J ) 7 Hz), 21.86 (s),
1.68 (s), 21.53 (s), 20.23 (d, J ) 6 Hz), 19.87 (s), 19.65 (s), 19.50
8
, 25
°
(
2
(
31
1
s), 19.17 (d, J ) 5 Hz), 18.76 (s), 18.65 (s). P{ H} NMR (126
MHz, THF-d , 25 °C): δ 69.06 (br s), 59.03 (br s). Anal. Calcd
found) for C34 Ni S: C, 52.61 (52.67); H, 8.57 (8.02); N,
.61 (3.60).
Computational Details. When available, known experimental
8
(
H
66
N
2
2
P
4
JA104158H
3
(
(
(
(
26) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
structures for the complexes were used as the starting point for the
calculations. To simplify these calculations, the i-Pr groups were
substituted by methyl groups. This simplification is assumed to have
no steric outcome on the calculations, and in selected cases, full
optimization with the dippe ligand was performed. The gas-phase
27) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
28) Frisch, M. J.; et al. Gaussian03; Gaussian, Inc.: Wallingford, CT, 2004.
29) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Hollwarth, A.;
Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 111.
25
structures were fully optimized in redundant internal coordinates,
with density functional theory (DFT) and a wave function incor-
(30) Hollwarth, A.; Bohme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi, A.;
Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 237.
(
31) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
(
25) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput. Chem.
2257.
1
996, 17, 49.
(32) Barone, V.; Cossi, M.; Tomasi, J. J. Chem. Phys. 1997, 107, 3210.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 35, 2010 12421